EX 1: Field Methods - Creating a Varied Elevation Surface

Introduction:

This activity was a group assignment to build a variable surface of each individual group's design. The activity introduced students to many elements/methods of interpolation. In addition to the previously stated methods, students tapped into their critical thinking skills to effectively model their surface, predict future weaknesses associated with their data collection scheme, and analyze their data for additional improvement.

Because human error is a part of every study, picking the best interpolation method on a group basis, and explaining why that method is best, will be fundamental to the learning and understanding to the assignment. The ultimate goal of the survey will be to further the understanding of the numerous survey techniques and enhance the critical thinking and creativity of students for future projects.

Exercise Requirements:
Sandbox Terrain must include: a ridge, hill, valley, depression, and plain.

Methods:

Given a two week window to finish the assignment and the uncharacteristically long cold snap, ideal weather conditions were not forthcoming. So, using weather underground.com the most favorable day of temperature, precipitation, and wind speed was chosen (7 °F/-14°C).

Since the ground was solid snow and sand, we had to break up the surface. Using both sides of the hammer interchangeably I broke up the hardest areas of the plot. Meanwhile the other group members used their hands or a Tupperware with a sharp lip to scoop and mold the surface.

Figure 1. With varied densities of the snow, ice, sand, and dirt.
Many tools were used in molding the surface of the sandbox.
Some of the tools used were: a hammer, sharp lipped tupperware,
 metal scraper, meter stick, and hands.

Figure 2. A squirt bottle filled with water was evenly
applied to the entire surface. Next regions with more
topographical variation were focused on to create
a more solid surface. 

Once we were satisfied with topography of the sandbox, I used a squirt bottle with water to wet down the surface. Looking at past blogger posts, by students from last year's field methods class, I had discovered that the frozen surface was much easier to collect measurements from. This was because the frozen surface would allow you to rest/touch the surface with the measuring device, in our case it was a meter stick. Without the frozen surface you would have had to hold the meter stick over the soft snow, which would increase the chance of systematic error for each measurement, with the likelihood increasing as time went on and fatigue/concentration set in. Even with the frozen snow firmly in place, I still noticed my arms getting sore having to reach out across the planters box for the middle cells.

Figure 3. Using rope and twine we made a grid of the planters box. First marking off the 3" segments, then hammering in tacks to anchor the twine in place. Looking at past students we knew tape would not stick during the cold weather months and tack/nails was the best option.

Once a measured grid was created, we used a two man system to record the data. One person would record the distance from the surface to the twine at each intersection (in reference to the rows and columns of twine).

Figure 4. Facing a trade off between accuracy and time, we decided to make our own coordinate system in 3x3 inch squares. Then, using a meter stick, at the intersection of those grid-lines measured each square's depth in centimeters at the top right corner of each box.

Discussion:

As many Midwestern Americans have previously discovered, the cold complicates things. This project was no exception and this group did their best to overcome the obstacles presented. As I was the most acclimated to the current weather conditions, I was able to stay outside with exposed hands the longest. This helped a lot because fastening the twine to the small tacks was difficult and time consuming when wearing cumbersome gloves/mittens. To overcome the cold group took shelter inside the nearby science building to warm-up whenever they needed a break from the cold. The use of a hammer was also a helpful tool, as it sped up the tacking of the twine and didn't require as much finger dexterity.

The hammer also helped in breaking up the frozen surfaces which would have otherwise been impenetrable. Using both sides interchangeably, the hammer was able to break up the ice as well as carve deep into the frozen sand. The Tupperware was also effective at excavating snow that would have otherwise taken increased energy to remove; however, it lacked the leverage required to break through the most dense regions of sand and/or ice.

To better represent the surface of the plot box, it would have helped to collect points along the ridges. This would have help preserve the flat nature at the top of the model. Instead our interpolated models showed peaks and ridges where in fact there were plateaus.

It would have been better to run the models earlier in the week. Then I would have saw the anomalies in ArcScene and resampled that one area with the spike.

Obstacles encountered during exercise one.

Conclusion:

This portion of the topology modeling was more about planning and group cohesion. Since the interpolation and computer models were associated with the next lab, knowing the parameters and inputs were a necessity for this lab. We had to ask ourselves what would be an effective technique for gathering data? Or what features would be difficult to model? Throughout all of this we had to contend with the temperature, precipitation, and wind. With teamwork we were able to work in shifts and use the frigid temperatures to our advantage.

If I was to perform this exercise again in the future I would try to isolate the sandbox from the weather and collect data with a variety of data gathering techniques. Such methods as random sampling, cluster sampling the highly variable areas, or combining the systematic grid with cluster sampling. These could then be compared to each other. No matter what method of data collection that is used, it will always come down to the amount of time spent collecting data and precision. With more data you will have more precise and accurate models, its just a matter of how much time you want to spend collecting data.